Deng and Xu Hereditas (2018) 155:1 DOI 10.1186/s41065-017-0036-2 REVIEW Open Access Adaptation of human skin color in various populations Lian Deng1,2 and Shuhua Xu1,2,3,4* Abstract Background: Skin color is a well-recognized adaptive trait and has been studied extensively in humans. Understanding the genetic basis of adaptation of skin color in various populations has many implications in human evolution and medicine. Discussion: Impressive progress has been made recently to identify genes associated with skin color variation in a wide range of geographical and temporal populations. In this review, we discuss what is currently known about the genetics of skin color variation. We enumerated several cases of skin color adaptation in global modern humans and archaic hominins, and illustrated why, when, and how skin color adaptation occurred in different populations. Finally, we provided a summary of the candidate loci associated with pigmentation, which could be a valuable reference for further evolutionary and medical studies. Conclusion: Previous studies generally indicated a complex genetic mechanism underlying the skin color variation, expanding our understanding of the role of population demographic history and natural selection in shaping genetic and phenotypic diversity in humans. Future work is needed to dissect the genetic architecture of skin color adaptation in numerous ethnic minority groups around the world, which remains relatively obscure compared with that of major continental groups, and to unravel the exact genetic basis of skin color adaptation. Keywords: Skin color, Natural selection, Genetic adaptation, Modern humans, Archaic hominin Background forms, pheomelanin (yellow-reddish) and eumelanin Since modern humans ventured out of Africa ~100,000 years (black-brown). The former is mainly accumulated in the ago, they spread across continents into a variety of habitats, light-complexioned people, while the latter is mostly from tropical zones to the arctic, and from lowlands produced in the dark-complexioned people [1–5]. In to highlands. During migration, selective pressures in addition, the number and size of melanin particles differ local environments (e.g., the cold climate, hypoxia, among individuals, and is even more important than the and endemic pathogens), together with random drift, proportions of the two forms of melanin in the deter- have resulted in population-specific genetic variants, mination of human skin color [5]. Other skin-related which further influenced variable phenotypes, such as lac- factors, e.g., keratin, also contribute to skin color tose tolerance, height, immune system, and metabolic variation [6, 7]. efficiency. In global populations, skin color is highly correlated Skin color variation is one of the most striking exam- with latitude, and fundamentally, the distribution of ples of human phenotypic diversity. It is dominated by ultraviolet (UV) radiation (Fig. 1). Populations closer to melanin, a pigmentation located in the base of the epi- the equator tend to have dark skin for protection against dermis and produced by melanocytes. Melanin has two UV, since overexposure to UV may decrease folic acid levels [8, 9] and cause skin cancer [10–13]. The lighter * Correspondence: [email protected] skin in populations at higher latitudes is underlying 1Chinese Academy of Sciences (CAS) Key Laboratory of Computational Biology, Max Planck Independent Research Group on Population Genomics, selection to maintain vitamin D photosynthesis, which is CAS-MPG Partner Institute for Computational Biology (PICB), Shanghai a UV-dependent process [14, 15]. Institutes for Biological Sciences, CAS, Shanghai 200031, China Although UV has been assumed to be a driving force 2University of Chinese Academy of Sciences, Beijing 100049, China Full list of author information is available at the end of the article for the evolution of human skin colors, understanding © The Author(s). 2017 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Deng and Xu Hereditas (2018) 155:1 Page 2 of 12 Fig. 1 Correlation between skin color and latitude (from Barsh (2003) [5]). (a) A map of human skin color distribution. (b) A plot of skin reflectance against latitude the exact genetic mechanism of selection would be gene identified in European is MC1R [29–31]. This gene is crucial to reconstruct human evolutionary history and expressed in melanocytes and plays a key role in control- elucidate the microevolution of adaptive traits. Describ- ling the switch from pheomelanin to eumelanin [31]. The ing a full picture of regional skin color adaptation in pigmentary phenotypes associated with MC1R has been humans would be challenging because it includes not studied in a wide range of animals [32–34]. Many variants only the genes identified to be under selection, but also have been identified in MC1R, such as rs1805007, the extent to which these genes could explain pheno- rs1805008, and rs3212357 [35, 36], despite its small size typic variation, the interactions and joint effects of (951 bp). Other important European-specific loci include genes, and the way they react to the external environ- rs1393350 in TYR, rs2733831 in TYRP1, and rs1900758 in ments. In this article, we reviewed several cases of skin OCA2 [17, 28, 37–39]. The derived allele frequencies at color adaptation in various populations of modern these loci are high in Europeans but low in Africans and humans and archaic hominins. These cases show the East Asians, which could be a clear signal of positive selec- similarities and differences of mechanisms of skin color tion in Europeans, as indicated by statistical analysis [40]. adaptation across populations, and provide some insights Genes involved in the skin color adaptation in East into human evolutionary history. Asians are not that well studied compared to the long list of adaptive genes identified in Europeans. Notable Skin color adaptation in modern Eurasians examples include OCA2 and MC1R. Each harbors sev- In Europeans, SLC24A5 and SLC45A2 [16–19] are two eral non-synonymous mutations, e.g. rs1800414 and golden genes related to the evolution of the light skin rs74653330 in OCA2, and rs885479 in MC1R [40–43], color. SLC24A5 encodes the NCKX5 protein, which is a which exhibit high derived allele frequencies in East member of the transmembrane protein family and regu- Asians, but low derived allele frequencies in Europeans lates the calcium concentration in the melanosome [16]. and Africans (Fig. 2). The OCA2 protein is thought to This gene has been confirmed to affect pigmentation in be a mature melanosomal membrane protein [44], with zebrafish and mice [16, 20]. Especially, the derived allele a potential role in protein transportation into melano- of rs1426654 in SLC24A5 was found to be nearly fixed somes [45]. The East Asian-specific variant of rs1800414 in Europeans, but almost missing in populations without was first reported in an exome sequencing study aiming any European ancestry (Fig. 2) [21]. A 78-kb haplotype to figure out albinism-related variants [46]. The derived around SLC24A5, which is in linkage disequilibrium with allele at rs1800414 was thought to contribute to the skin rs1426654, was also identified to accumulate in Europeans lightening in an association study of Han Chinese, which [22]. A similar pattern can be observed at rs16891982 measured the skin color of individuals using the melanin in SLC45A2 [23], which has been reported to be asso- index [47]. Another non-synonymous variant in OCA2, ciated with pigmentation in several species, e.g., mice, rs74653330, has also been confirmed to be pigmentation- fish,birds,andhorses[24–26]. Other variants in this related in an association study of Japanese [48]. Additional gene, including rs26722, rs2287949, and rs40132, were examples of East Asian-specific pigmentation-associated also shown to be coloration-associated in Europeans alleles include rs10809814 in TYRP1 and rs1407995 in [23, 27, 28]. Another important pigmentation-related DCT [40, 49], both of which show differentiation between Deng and Xu Hereditas (2018) 155:1 Page 3 of 12 Fig. 2 Evolutionary model of human pigmentation in three continental populations. The rooted tree shows the genetic phylogeny of human populations from Africa, North Europe and East Asia, with the colors of the branches roughly indicating the generalized skin pigmetation level of these populations (adapted from McEvoy et al. (2006) [39]). Genetic loci reported to be under positive selection in the common ancestor of modern Eurasians are represented by rs1881227 in KITLG, and those independently evolved in Europeans and East Asians, indicating possible convergent evolution, are represented by rs12913832 in OCA2 and rs885479 in MC1R, respectively. The maps of allele frequency were drawn using R (version 3.2.1, https://www.r-project.org), based on these loci in 53 global populations provided by the Human Genome Diversity Panel CEPH (HGDP, http:// www.hagsc.org/hgdp/index.html). Blue and red colors denote the ancestral and derived alleles, respectively Asians and non-Asians [47], and strong signals of positive is located at 326 kb upstream to the transcription start selections in Asians [43, 49]. site of KITLG. At this variant, the ancestral allele Despite distinct genes and variants under respective frequency is over 90% in Africans, comparable to the local adaptations in Europeans and East Asians, some derived allele frequency in Europeans and East Asians genes have derived alleles reaching high frequencies in (Fig. 2). Similar patterns were observed in other genes, both continental groups. For instance, KITLG exhibits a e.g., ASIP and BNC2 [39].